This skill should be used when the user asks to "segment an image", "identify objects", "extract objects", "generate masks", "find objects in image", or needs AI-powered image segmentation.
From gemskillsnpx claudepluginhub b-open-io/claude-plugins --plugin gemskillsThis skill uses the workspace's default tool permissions.
scripts/segment.tsDispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Executes pre-written implementation plans: critically reviews, follows bite-sized steps exactly, runs verifications, tracks progress with checkpoints, uses git worktrees, stops on blockers.
Guides idea refinement into designs: explores context, asks questions one-by-one, proposes approaches, presents sections for approval, writes/review specs before coding.
Segment and identify objects in images using Gemini's vision capabilities.
Use this skill when the user asks to:
bun run --cwd ${CLAUDE_PLUGIN_ROOT} ${CLAUDE_PLUGIN_ROOT}/skills/segment-image/scripts/segment.ts <input-image> [options]
--prompt <text> - Custom segmentation prompt--output <dir> - Output directory for mask files# Segment all objects
bun run --cwd ${CLAUDE_PLUGIN_ROOT} ${CLAUDE_PLUGIN_ROOT}/skills/segment-image/scripts/segment.ts photo.jpg
# Segment with custom prompt
bun run --cwd ${CLAUDE_PLUGIN_ROOT} ${CLAUDE_PLUGIN_ROOT}/skills/segment-image/scripts/segment.ts photo.jpg --prompt "identify all people and vehicles"
# Save masks to directory
bun run --cwd ${CLAUDE_PLUGIN_ROOT} ${CLAUDE_PLUGIN_ROOT}/skills/segment-image/scripts/segment.ts photo.jpg --output ./masks
Do not read generated mask images back into context. The script outputs file paths. Ask the user to visually inspect the masks. To inspect programmatically, optimize the images first (via the optimize-images skill).
Uses gemini-3-flash-preview (Gemini 3 Flash) for image segmentation.
Last verified: February 2026. If a newer generation exists, STOP and suggest a PR to
b-open-io/gemskills. See the ask-gemini skill'sreferences/gemini-api.mdfor current models and Google's officialgemini-api-devskill for the canonical source.